{ datagubbe }


datagubbe.se » the future of the software industry

The Future of the Software Industry

On geopolitics, economics, class, looming austerity... and Twitter.

Summer 2023

I've previously written about the possible future of FOSS software and, in many of my texts, mumbled about the bleak future of the software business in general. I've thought about writing about this for a long time, but for various reasons deemed it uninteresting. Yet, I keep ruminating on it. Here are some unsorted thoughts, loosely linked together to form a whole. These are guesswork theories at best, but nevertheless, it seems a pattern is emerging from which conclusions might be drawn.

The three tiers of software

Let us, for the sake of argument, divide software into three tiers. It may not be perfect, but still very usable to broadly define what types of software our industry produces - and what might happen to it during bad times.

Frivolous

The topmost tier, making up the bulk of the recent software boom, is completely frivolous software. This means everything from novelty and lifestyle apps to games to putting software in items that really don't need software, such as fridges and door locks. This is likely the first category of software to be prioritized away by consumers during hard times: things that are nice conveniences if you've got the cash to spare, but that you can honestly do without.

Frivolous software won't go away completely - even hard times call for entertainment. And it might not always be the consumer deprioritizing: many of these apps are made by startups or financed by ads, which means there's going to be a coming crunch to find other ways to monetize - or go bust. Reddit's recent API pricing changes is one example of this.

Bread and butter

This is the type of software most office workers are familiar with. Word and Excel are in this category, but many other types of software fall into this category as well. It might be unknown to the general public, but nevertheless of importance to most companies, ranging from behemoths like SAP to smaller, regional players that know the ins and outs of legislature or other local particularities regarding for example bookkeeping, taxation and logistics. When times are bad, the user base for this kind of software shrinks. Smaller companies may merge with bigger ones or otherwise struggle economically. The end result for programmers may be the same; less work to do.

Essential

Increasing amounts of infrastructure, both physical and financial, rely on software. As do societal necessities like hospitals, police forces and armies. A small core of software keeps these systems running and has, in many cases, been doing so since long before the home computer boom, the web and the smartphone. This software is likely to be the very last to suffer the effects of economic hardships, but it also makes up a relatively small amount of current programming jobs.

Recent tech layoffs

A lot of the recent layoffs in big tech seem to be linked to overhiring during the peak of the COVID-19 pandemic, and I think there's some truth to this. Nevertheless, this doesn't lessen the individual hardships of getting laid off. Neither does it inspire confidence in the overall management of these companies, nor do the mass layoffs end there. Take, for example, Elon Musk's recent acquisition of Twitter. Regardless of what one may think of it and him, staff has according to Musk himself been cut from roughly 8,000 to 1,500. Despite this, the site is still up and running and has even introduced some new features. Its position as prime hangout for politicians, policy wonks, pundits, activists and journalists seems unthreatened so far.

Too early to call? Perhaps. Yet, it does say something about the amount of staff needed to operate and develop one of the biggest social media platforms in the world. Being in the software business myself, I can of course understand why we're not too eager to listen.

Overproduction

I've previously written at length about the theory of elite overproduction. In short, elite overproduction is when a surplus of people with elite aspirations (E.G. middle class college graduates) meet an economy with little or no real use for their skills. Social unrest will ensue, unless they can somehow be incorporated into the workforce in a way that placates their ambitions. This isn't limited to software companies, but can be seen everywhere in western society as an increasing number of jobs very far removed from the actually productive core parts of businesses. In some cases, these are jobs likely to be cut first, and in some cases they're likely to be cut last.

Twitter, for example, was and is getting a lot of flak for firing staff that had little to do with the actual production of software, such as public policy and human rights teams. Views on whether this was a good or bad idea seem to align with a larger polarizing divide in the west.

This divide is perhaps best explained using a counterexample in the form of Mozilla, a curious construct that's both an NGO and a for-profit company. Despite laying off 200 developers during otherwise good times, CEO and chair Mitchell Baker (and the Mozilla Foundation) remains popular with the same crowd that's lamenting Twitter's supposed fall from decency. The key difference here is the core business: The Mozilla Foundation isn't really about making software, it's about producing whatever ideology that'll bring in the most status and money for its board of directors and their immediate staff. If the Mozilla Corporation went defunct tomorrow, the Foundation would just double down on its ideology production and keep passing the hat. And, for a growing group of people, remain a possible employer.

Nobody woke the plumbers

This mirrors a much larger economic conflict, the nature of which divides the tech business itself. On the one hand, the same kind of ideology producers that run the Mozilla Foundation are often the ones that decide what open source projects will get funded, the content moderation policies of major social media, what kind of computer games should get made and what type of personal conviction the employees of certain corporations are expected to hold. On the other hand, this type of activity funnels money away from programming, not just directly by way of salaries, but also in the form of externalities that cut into overall profitability and productivity.

Programmers are not unaffected by this divide. They may align with either side depending on their personal convictions, their ambitions and aspirations, and how their current employer is funded. I imagine someone involved in programming for offshore oil drilling systems is less concerned about the latest progressive talking points than, say, a VC funded Silicon Valley startup that wants to simplify organic grocery shopping. The same can be said about other professions as well: few people care if their plumber is "woke", but if you're working HR in AAA gaming, you have to be. The divide is, roughly speaking, about essential versus frivolous.

Marxmaxxing

Hence, a conflict is brewing here that cuts through traditional class divides and could, possibly, cause implosions not just in corporate leadership but the companies themselves. Is it more important for FAANG to adhere to preferred political standpoints and hire ideological commissars to enforce those, or do they in fact have a core business which must be kept running? The recent layoffs mentioned above seem to echo the latter, but the commissars are hardly left without power.

In the case of Twitter, Musk's other ventures are in many ways dependant on the benevolence of people he's now bought a social media platform to troll. Tesla, Tesla Energy, The Boring Company and SpaceX all rely on various subsidies and grants, and hence the goodwill of politicians who must also answer to the call of a sizeable progressive voter base. In Sweden, for example, all electrical cars have until recently been heavily subsidized by the government. Generous US tax grants for electrical vehicles are tightly tied to specific makes and models.

Flick those policymakers and their friends on the nose for too long, and see what happens.

On the other end of the corporate spectrum, we find the recent goings on with the Bud Light brand. A miscalculation about the consumer base, perhaps, led to a misstep in marketing. But don't think for a second that Alissa Heinerscheid is somehow corporate pariah - my guess is that a number of companies and organizations are going to line up if she decides to jump ship.

Within our current societal dynamics, going woke is hardly synonymous with going broke, and neither is anyone completely uncancellable. At the same time, profit is, after all, still the point of any corporation. A society will have a hard time busying itself with semantics if basic infrastructure remains broken for too long, and the fruits of worker exploitation can only be enjoyed if there is some actual work to be exploited. Thus, the scene is set for what ostensibly plays out as a culture war, but in fact boils down to the same old fight over resources it's always been.

Out of money error

Despite massive cuts in staff, profit is the one cloud on the skies of Twitter left to zap the bird with lightning. Like many tech companies, Twitter costs a lot of money to operate and looks pretty far from turning into a reliable cash cow. However, even its most fervent detractors don't seem to be speaking too much about this, but rather about how the tech platform itself will fail miserably due to said staff downsizing. Perhaps because talk about profitability would attract attention to the many ideologically palatable software companies (such as Spotify) that are still completely dependent on venture capital. Nobody wants to jinx the good times into ending.

Alas, end they do, and there should be little doubt that we're currently at the beginning of that end. US Treasury Secretary Janet Yellen recently conceded that the sanctions imposed on Russia could "over time" undermine the US dollar's hegemony as world reserve currency. Recent developments are surely pointing in that direction: one example is France buying Russian liquid natural gas via China and paying in Chinese renmibi. The BRICS nations are, as we're speaking, building their own financial infrastructure and deepening their collaboration with middle eastern OPEC nations. It's not so much about replacing a reserve currency; simply not using one will suffice.

This means quantitative easing will get harder for the Federal Reserve, further hampered by inflation that calls for interest rate hikes. The capital glut is coming to a close, one way or another, impacting the appetite for risk and access to funding for startups. One example of what this might look like is that Andreessen Horowitz is already courting Saudi Arabian capital.

In the world of ultra rich megacorps like Facebook, Google, Apple and Microsoft, this may seem unimportant, but bad times are bad for everyone. Parts of the FAANG are more exposed than others.

Facebook and Google, for example, are completely dependant on advertising. When times are bad, people buy less stuff. Instead of an increase in companies trying to peddle their product, this usually leads to slashed marketing budgets. Facebook is mostly in the business of producing frivolous software, whereas Google has a slightly more bread'n'butter tinge. Are they going away any time soon? Probably not, but I doubt we've yet seen the last massive staff cut at 1 Hacker Way.

European problems

If the sanctions against Russia are putting the US in an economically precarious position, this goes doubly for Europe. As vassals to the American empire, we're very sensitive to the whims of US policies and economy - but we also have problems completely of our own making. The previously mentioned fact that France is buying Russian LNG and paying in renmibi isn't merely a sign of dwindling dollar hegemony, it also signals that the sanctions imposed on Russia are impossible to deal with for the nations supposedly tasked with enforcing them.

This European energy crisis has only just begun. The recent winter passed relatively smoothly, but many more are coming - and politicians seem hellbent on making them worse. Nordstream remains blown up and at the time of writing, Germany is shutting down its last nuclear reactors. Several of the country's major industries have already made it abundantly clear that they cannot continue operations in Germany or even Europe, given the rising energy costs.

This will not only affect people's consumption patterns superficially, such as prioritizing away certain types of convenient phone apps. It's not even about mass unemployment across all sectors of the economy. It is, in the end, a question of whether citizens can expect continuous access to electricity at all. Politicians and media alike are already floating the concept of regular, rolling blackouts. Such events are usually not considered conducive to the use and production of software.

Venture Capital Ponzi vibes

Many ostensibly "tech" companies like Doordash, Airbnb and Uber are service brokers more than software companies, and whatever software they produce is completely frivolous. Even if they once dreamt of "disruptive" tech, such as self driving cars, those dreams have long since been replaced with trying to cut costs by subverting various labor regulations. More importantly, their customer base is the type of people who are now getting laid of en masse.

In fact, many of the hyped up VC funded "tech" companies have been propped up by their own hiring base: Tech biz adjacent upper middle class city dwellers accustomed to splurging on expensive gadgets, gimmicky phone apps, cab rides, vacations and food deliveries. In a sense, VC has been injected into a giant Ponzi scheme, making large parts of the software industry dependent not only on money for startups, but also on a constant supply of highly comped startup employees to which their own products and services have catered. As such, the layoffs are a strange kind of cannibalization, undermining the strength of the core consumer base. The question is if another route can be found or a balance struck somehow.

Even established companies can fall into this trap, such as Apple. Despite moving to custom silicon, they remain an expensive luxury brand offering little real edge for the bulk of developers and other office drones. In the end, having a fruit on your laptop is more about status and "lifestyle" than about productivity. Consequently, recent events show that they're not immune to layoffs, either.

Microsoft resides firmly in the bread and butter camp, with Windows forming a base even in many essential software systems. Much of their revenue these days is however coming from frivolous ventures, even if indirectly. Azure, as well as Google Cloud and AWS, are wildly successful because so much frivolous software is running in the cloud. In other words, the demand for this kind of digital infrastructure will likely shrink.

Unglobalization

With energy getting so expensive in Europe, wouldn't US cloud providers operating overseas become extremely competitive, bandwidth permitting? Perhaps to some extent - but most European authorities are rightfully wary of putting their eggs in a basket ultimately controlled by the FBI and NSA. This echoes the recent - and commendable - clamping down on Tiktok by many western nations. The Great Firewall of China is already a well known concept, but the EU and US are well on their way in constructing their own versions of it, accelerated by sanctions against Russia. The once sprawling, anarchic Internet is slowly but surely being tamed and sectioned off according to national interests.

Just as with FOSS, the global market for proprietary software is shifting with the geopolitical landscape. Multipolarity means the US will no longer be able to enforce intellectual property laws on a global scale. A dispute over Taiwan will make matters even worse, regardless of whether the US is even capable of getting directly involved in an armed conflict with China. Taiwan is the world's chip fab, an industry heavily dependent on imports. A trade blockade may not be as immediately devastating as all-out war, but even in our day and age, no software runs without hardware.

Bringing such industries "back home" to a Europe unable to sustain even its current dwindling industrial base is highly unlikely. The US also suffers from low industrial capacity even in critical sectors, incapable of replenishing its domestic arms stockpile in a timely fashion after large donations to Ukraine. US president Biden's current plan to reshore industry is commendable, but will take heroic measures of time and effort to succeed. In Europe, leaders seem uninterested in even talking about doing something similar.

Looking back - and forward

The software business has, despite dotcom busts and various other financial crises, only ever experienced good times: cheap energy, rising empire and - for more than thirty years - unipolar US hegemony. It's hardly a high stakes gamble to suggest that businesses will do worse in times of economic decline, but the dynamics currently affecting our industry are completely new to it and us. Conflicts will likely appear along highly confusing lines, dividing not just the industry but also the internal workings of many corporations.

Software is here to stay, in one form or another, but a decades long party is undoubtedly coming to a close. A slow decline seems much more likely than a quick rebound. However one extrapolates, the future is sure to be - at the very least - humbling.